Unidad Didáctica

ANALYSIS OF Δ

_{-TETRAHYDROCANNABINOL AND ITS TWO MAIN}

METABOLITES IN WHOLE BLOOD USING AUTOMATED DISPERSIVE PIPETTE

EXTRACTION AND LC-MS/MS

Abstract

An analytical procedure was developed and validated for the analysis of Δ9_-

tetrahydrocannabinol (THC) and its metabolites (11-hydroxy-Δ9-tetrahydrocannabinol (11-OH-THC) and 11-nor-9-carboxy- Δ9-tetrahydrocannabinol (THC-COOH)) in whole blood using LC-MS/MS. An automated dispersive pipette extraction (DPX) using a unique liquid-liquid solid-phase extraction was employed on a Hamilton NIMBUS96 platform to extract the analytes of interest. Extraction time was less than 3 min with a total LC-MS/MS run time of 5.6 min. The method was fully validated in accordance with the Scientific Working Group of Forensic Toxicology (SWGTOX) guidelines for limit of detection (0.22 ng/mL THC, 0.25 ng/mL 11-OH-THC and 0.62 ng/mL THC-COOH), limit of quantitation, carryover, extraction efficiency (93-100%), matrix effects (8-30%), linearity (0.5-50 ng/mL), within and between-run precision (CV <7.5%), and accuracy (mean relative bias <5%). An interlaboratory comparison of patient samples in

collaboration with a local forensic toxicology lab method resulted in a correlation coefficient of 0.9901 between results from the two labs.

Introduction

Cannabis is the most widely abused illicit drug in the US. According to the National Survey on Drug Use and Health, 18.9 million people admitted to marijuana use in the previous month. Between 2007 and 2012, the rate of marijuana use rose from 5.8% to 7.3% (1). As the use of marijuana increases, it becomes more prevalent in clinical and forensic case work, particularly in impaired driving cases. After legalization in

Washington State in 2015, the percent of cannabinoid positive cases went from 28 to 40% (2). Colorado also reported an increase in cannabinoid positive case samples after legalization in 2012 (3). The National Highway Traffic Safety Administration (NHTSA) reported in 2007, that 8.6% of nighttime drivers tested positive for cannabinoids  a rate almost four times higher than those with blood alcohol levels equal to or above 0.8 g/L (4). Recent use of marijuana is associated with 2-6 times increased risk of crashing while driving, depending on dose, than when unimpaired (5). From 1992 to 2009, 20,000 US drivers involved in fatal car crashes tested positive for cannabinoids (5).

Δ9_{-Tetrahydrocannabinol (THC) is the main psychoactive ingredient in cannabis} (marijuana). After smoking, THC is rapidly absorbed into the blood stream. THC is metabolized into two main metabolites, the active metabolite, 11-hydroxy-Δ9_-

tetrahydrocannabinol (11-OH-THC) and the inactive metabolite, 11-nor-9-carboxy- Δ9- tetrahydrocannabinol (THC-COOH). THC affects mental processes, ranging from altered perception of time and distance to hallucinations (6), owing to the risk of use while driving. Peak THC effects are 20-30 min after use, when blood levels are the highest, levels are low after 3 h and become baseline after 4 h (7, 8). Blood is routinely the matrix of choice when determining drug or alcohol impairment. Colorado and Washington states

have both legalized recreational marijuana use and adopted a 5 ng/mL THC blood concentration as the driving under the influence of drugs (DUID) level (3, 9).

Unfortunately, blood draws are not done at the “scene” and can take up to hours after the time of the incident. DUI data shows that 42 and 70% of all cannabinoid-positive traffic arrests tested below 5 ng/mL THC in blood. Detection and quantification of THC and its metabolites in blood at low levels (< 5ng/mL) are imperative in determining time of THC use and potential impairment.

Previously published LC-MS/MS methods for the analysis of THC and its metabolites in blood require at least 0.25 mL of blood (10), but most require 0.5 mL or more (6, 11-16) or derivatization (17) to achieve necessary sensitivity. These methods often necessitate tedious sample preparation, e.g., use of SPE columns (11-13, 15-16), intricate online SPE (6), or liquid-liquid extraction (10, 14). Our aim was to develop a method that minimizes sample volume and automates a fast and easy dispersive pipette extraction procedure to obtain sensitive quantitation of THC, 11-OH-THC and THC- COOH in whole blood using LC-MS/MS.

Experimental Methods

Reagents and Standards.

All drug standards (Δ9-Tetrahydrocannabinol, 11-hydroxy-Δ9-tetrahydrocannabinol, 11- nor-9-carboxy- Δ9-tetrahydrocannabinol, Δ9-Tetrahydrocannabinol-d3, and 11-nor-9- carboxy- Δ9-tetrahydrocannabinol-d3) were purchased from Cerilliant Corporation (Round Rock, TX). DPX WAX-S tips were purchased from DPX Technologies, LLC (Columbia, SC).

Instrumental Analysis.

Analyses were performed using a Thermo TSQ VantageTM triple quadrupole mass spectrometer (Milwaukee, WI) coupled to an Agilent 1260 Series HPLC (Agilent

Technologies, Santa Clara, CA) equipped with an Agilent Poroshell EC-C18 column (3.0 × 50 mm, 2.7 µm) with column temperature held at 50 °C. Sample injections of 20 µL were made using a 6 port (0.25 mm) Cheminert C2V injection valve (Houston, TX) incorporated on a dual rail GERSTEL MPS autosampler (Linthicum, MD).

The mobile phase was composed of 0.1% formic acid in water (A) and 0.1% formic acid in methanol (B). The gradient started at 40% B, ramped to 90% B at 1.3 min, and 91% B at 3.5 min, after which the composition was ramped quickly ramped to 98% B at 3.6 min, where it remained until 4.6 min and was re-equilibrated to 40% B. The total run time was 5.6 min. Eluent was diverted to waste during the intervals of 0-1.5 min after injection. The column flow rate was 0.5 mL/min. Mass spectrometer parameters were: electrospray voltage, 5000 V/-4500V; auxillary gas pressure, 25 psi; sheath gas pressure, 35 psi; vaporizer temperature was 330 °C, and capillary temperature was 400 °C. Ion transitions monitored for each compound are listed in Table 4.1.

Sample Preparation.

Aliquots of 100 L of each sample (calibrator, control, blank, patient sample) were transferred to a 2 mL micro centrifuge tube (VWR, Radnor, PA). Internal standard in methanol was added (10 µL) and the tubes were vortex mixed. Acetonitrile (300 µL) was added and the tubes were vortexed using a Fisher Scientific (Waltham, MA) Vortex- Genie at speed 10 and then centrifuged for 10 min at 13,300 RPMs using a Thermo

Scientific Sorvall Legend Micro 17 centrifuge (Milwaukee, WI). For the interlaboratory comparsion, the forensic samples were prepared up to protein precipitation at the State Law Enforcement Division to avoid any biological hazards at the University of South Carolina. The protein precipitated forensic samples were brought back to the university for centrifugation and further analysis. The supernatant (350 µL) was transferred to a 2.2 mL well plate, which was then placed on the Hamilton NIMBUS96 system (Reno, NV) for the automated solid-phase extraction procedure. The NIMBUS system was loaded with DPX WAX-S tips (10 mg WAX resin and 20 mg salt), 300 µL CO-RE tips, a reservoir of 0.1 M formic acid, and an additional empty well plate. The NIMBUS system uses the CO-RE tips to add 50 µL of 0.1 M formic acid to the well plate containing the sample supernatant. The 1 mL DPX WAX-S tips are then used to aspirate and dispense the sample solution three times, thus allowing extraction of matrix and subsequent partitioning of the acetonitrile and the aqueous phase. The CO-RE tips are used to transfer 100 µL of the supernatant (acetonitrile layer) to a clean well plate, which is then transferred to the LC-MS/MS for injection.

Method Validation

Linearity and Sensitivity.

The method was validated according to SWGTOX guidelines (18). Linear least squares regression with a 1/x weighting was used as a calibration model for each analyte and spanned 0.5 ng/mL to 50 ng/mL with seven calibration points and five replicates at each point. Carryover was evaluated by running an extracted blank matrix sample after each high calibrator (n = 5). The sensitivity was determined by calculating the limits of

detection and quantitation. The limit of detection (LOD) was calculated using the standard deviation of the y-intercept (sy) and the average slope (avgm):

LOD = (3.3 sy)/avgm

The limit of quantitation (LOQ) was similarly quantified with a multiple of 10 instead of 3.3. The average slope and standard deviation of the y-intercept were taken from the five separate runs of the calibration plot.

Accuracy and Precision.

The accuracy and precision of the method were determined by evaluating three different concentrations (2 ng/mL, 10 ng/mL, and 25 ng/mL) in triplicate over 5 separate runs. The accuracy was calculated as the (mean concentration measured – fortified concentration) divided by the fortified concentration × 100%. The within-run precision was determined by taking the standard deviation of a single run of samples at a single concentration divided by the mean calculated value of that single run × 100%. The between-run precision was determined by taking the standard deviation of all observations for each concentration divided by the grand mean for each concentration × 100%.

Extraction Efficiency and Matrix Effects.

Matrix effects were determined using the post-extraction addition technique. An

unextracted neat solution of a low concentration (1 ng/mL 11-OH-THC, 5 ng/mL THC, and 10 ng/mL THC-COOH) and a high concentration (2 ng/mL 11-OH-THC, 20 ng/mL THC, and 40 ng/mL THC-COOH) were injected six times each (set 1). These results were compared to pooled blank blood spiked with analyte at the appropriate

concentration post-extraction in triplicate (set 2). Matrix effects were then calculated as the mean area of set 2 divided by the mean area of set 1 subtracted from 1 and multiplied by 100%. A negative value represents ion suppression, while a positive value represents ion enhancement. Extraction efficiency was determined by comparing the matrix

matched samples (set 2) to a set of samples that were fortified after protein-precipitation and centrifugation, but before extraction (set 3). Extraction efficiency was calculated as the mean area of set 3 divided by the mean area of set 2 multiplied by 100%. Finally, loss during the protein precipitation step was also determined. In this case, the post-

precipitation spiked samples (set 3) were compared to a set of samples where the blood was fortified before any processing (set 4). Protein precipitation loss was also calculated as the mean area of set 4 divided by the mean area of set 3 × 100%.

Results and Discussion

LC-MS/MS parameters.

Various liquid chromatography and mass spectrometry parameters were evaluated, but the most noteworthy differences involved the spray voltage and the choice of organic mobile phase. The difference in analyte signal when switching from low (3500/-3500V) to high (5000/-4500V) spray voltage was significant. The 11-THC-OH analyte showed the largest difference with a 2.5-fold increase in signal intensity after switching from low to high voltage. Similarly, the THC peak doubled in intensity, but the THC-COOH peak showed a 33% increase. The change of organic mobile phase (B) from acetonitrile to methanol also resulted in significant increases in signal intensity for all analytes. The 11- THC-OH peak exhibited the largest signal increase of over 8 ×. The THC signal

increased by almost 4 times, and THC-COOH signal increased by about 2.5 times. The reduced surface tension of methanol and the increased spray voltage both helped to increase ionization efficiency, with a concomitant increase in both signal intensity and sensitivity.

Stability in Plastic Time Study.

It is well known that THC has a tendency to fall out of solution and/or be retained on the surfaces of its container, particularly plastic. In the method presented here, plastic well plates were utilized throughout sample preparation. Most importantly, plastic well plates hold the eluent during LC-MS/MS analysis. With a run time of 5.6 min, a full well plate could take 9 h. A study was performed to evaluate analyte loss due to the use of plastic well plate containers at room temperature in the eluent conditions (~100% acetonitrile). Samples were extracted in triplicate at three different concentrations (1, 10, 25 ng/mL) for each analyte. The extracted samples were injected immediately at time 0, and then again at 6, 12, and 24 h. Concentrations did not vary by more than 15% except at 1 ng/mL for THC-COOH, which decreased by 35% after 12 h; however, no further loss was observed at 24 h. As a result of this study, well plates were used throughout this study, but calibrators were re-injected at the end of the runs to ensure stability.

Linearity and Sensitivity.

Calibration fitting resulted in average coefficients of determination (R2) values of 0.9984 for THC, 0.9980 for 11-OH-THC, and 0.9966 for THC-COOH. The average slope and y- intercept standard deviation values were used to determine the LODs and LOQs for each compound (0.66 ng/mL for THC, 0.75 ng/mL for 11-OH-THC, and 1.8 ng/mL for THC-

COOH) as described above (Table 4.2). The limit of quantitation for each compound was well below the recommended cut-off level for DUID confirmation in blood of 1, 5, and 1 ng/mL for THC, THC-COOH, and 11-OH-THC, respectively (19). Chromatograms of the parent-to-quantifier transition response for each of the analytes at the suggested cut- off levels are shown in Figure 4.1.

Accuracy and Precision.

Accuracy, within-run precision, and between-run precision were determined and shown in Table 4.3. The method exhibited a minimum bias of 0.02% at 25 ng/mL for 11-OH- THC, and a maximum of 4.7% at 2 ng/mL of THC. The within-run and between-run precision was ascertained from the same replicate analyses. The average within-run precision had a maximum of 6.8% at 2 ng/mL of THC, and the between-run precision had a maximum at 2 ng/mL of THC at 7.5%.

Extraction Efficiency and Matrix Effects.

Low concentrations of analyte are most susceptible to large matrix effects. Ion

suppression for THC was 31% at 5 ng/mL, but only 12% at 20 ng/mL. The 11-OH-THC analyte elicited ion suppression at 1 ng/mL of 6% and 2% at 2 ng/mL. THC-COOH had negligible matrix effects, showing ion suppression at 10 ng/mL of 3% but ion

enhancement at 40 ng/mL of 3%. The DPX extraction post-protein precipitation was very efficient at >93% at both concentrations for all compounds. The loss of analyte during the protein precipitation step also varied with concentration. There was approximately 30% loss of all compounds at 25 ng/mL, but 15%, 19%, and 8% of THC, 11-OH-THC and THC-COOH, respectively, were lost at 2 ng/mL.

Interlaboratory Comparison.

A successful sample comparison was completed with the South Carolina Law

Enforcement Division (SLED). SLED uses a method that requires 1 mL of blood and employs solid-phase extraction cartridges and GC-MS analysis. Twenty-eight forensic samples were compared. Each sample was analyzed in triplicate using the DPX method described above. The coefficient of determination (R2) for the comparison of 11-OH- THC positive samples was 0.9965 (n = 3) (Figure 4.2), THC-COOH R2 was 0.9836 (n = 27) (Figure 4.3), and THC produced an R2 of 0.9960 (n = 16) (Figure 4.4). When all positive results were combined, the overall correlation was 0.9901 (Figure 4.5). A

chromatogram of the parent to quantifier ion transitions for each of the analytes in patient sample 13 is shown in Figure 4.6. The calculated percent difference in triplicate analysis of each sample compared to the reported concentration from SLED did not exceed 20% for any case sample. The relative standard deviation of the triplicate extractions of patient samples did not exceed 15%. SLED detection cut-off values were 2 ng/mL for each analyte. Our LOQs were considerably lower and resulted in more positive 11-OH-THC and THC samples than SLED identified.

Conclusion

The LC-MS/MS method developed in this study minimizes required sample volume, and provides sensitive quantitation of THC, 11-OH-THC and THC-COOH in whole blood. Notably, our method simplifies sample preparation for the analysis of THC and its metabolites in blood with an automated dispersive pipette extraction without subsequent dilution or solvent evaporation. The extraction process is rapid, minimizes matrix effects

(<30%), and maximizes recoveries (>93%). LODs and LOQs were below 0.75 ng/mL and 2 ng/mL, respectively. These outcomes clearly provide the necessary sensitivity to meet laboratory cut-off with minimal imprecision (<8%). All calibrations were linear (R2 > 0.99) over two orders of magnitude (0.5-50 ng/mL). Lastly, a successful comparison of forensic case sample using our new method with a local toxicology laboratory verifies the effectiveness of this new quick and easy method.

References

1. Substance Abuse and Mental Health Services Administration, Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings, NSDUH Series H-48, HHS Publication No. (SMA) 14-4863. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2014.

2. Couper, F., Peterson, B. (2014) The prevalence of marijuana in suspected

impaired driving cases in Washington state. Journal of Analytical Toxicology, 38, 569-574.

3. Urfer, S., Morton, J., Beall, V., Feldmann, J., Gunesch, J. (2014) Analysis of Δ9- tetrahydrocannabinol Driving Under the Influence of Drugs Cases in Colorado from January 2011 to February 2014. Journal of Analytical Toxicology, 38, 575- 581.

4. Compton, R., Berning, A. (2009) Results of the 2007 national roadside survey of alcohol and drug use by drivers. COT HS 811 175. NHTSA, Washington, DC. 5. Masten, S., Guenzburger, G. (2014) Changes in driver cannabinoid prevalence in

12 U.S. states after implementing medical marijuana laws. Journal of Safety

Research, 50, 35-52.

6. Jagerdeo, E., Schaff, J., Montgomery, M., LeBeau, M. (2009) A semi-automated solid-phase extraction liquid chromatography/tandem mass spectrometry method for the analysis of tetrahydrocannabinol and metabolites in whole blood. Rapid

Communications in Mass Spectrometry, 23, 2697-2705.

7. Huestis, M.A. (2007) Human cannabinoid pharmacokinetics. Chemistry and

8. Grotenhermen, F. (2003) Pharmacokinetics and pharmacodynamics of cannabinoids. Clinical Pharmacokinetics, 42, 327–360.

9. Wood, E., Brooks-Russell, A., Drum, P. (2016) Delays in DUI blood testing: Impact on cannabis DUI assessments. Traffic Injury Prevention, 17, 105-108. 10.Fernandez, M., Boeck, G., Wood, M., Lopez-Rivadulla, M., Samyn, N. (2008)

Simultaneous analysis of THC and its metabolites in blood using liquid

chromatography-tandem mass spectrometry. Journal of Chromatography B, 875, 465-470.

11.Coulter, C., Miller, E., Crompton, K., Moore, C. (2008) Tetrahydrocannabinol and two of its metabolites in whole blood using liquid chromatography-tandem mass spectrometry. Journal of Analytical Toxicology, 32, 653-658.

12.Schwope, D., Scheidweiler, K. (2011) Direct quantification of cannabinoids and cannabinoid glucuronides in whole blood by liquid chromatography-tandem mass spectrometry. Analytical and Bioanalytical Chemistry, 401, 1273-1283.

13.Elian, A., Hackett, J. (2009) Solid-phase extraction and analysis of THC and Carboxy-THC from whole blood using a novel fluorinated solid-phase extraction sorbent and fast liquid chromatography-tandem mass spectrometry. Journal of

Analytical Toxicology, 33, 461-468.

14.Fabritius, M. et al. (2013) Comparison of cannabinoid concentrations in oral fluid and whole blood between occasional and regular cannabis smokers prior to and after smoking a cannabis joint. Analytical and Bioanalytical Chemistry, 405, 9791-9803.

15.Jamey, C., Szwarc, E., Tracqui, A., Ludes, B., (2008) Determination of cannabinoids in whole blood by UPLC-MS-MS. Journal of Analytical

Toxicology, 32, 349-354.

16.Simoes, S., Ajenjo, A., Dias, M. (2011) Qualitative and quantitative analysis of THC, 11-hydroxy-THC, and 11-nor-9-carboxy-THC in whole blood by ultra- performance liquid chromatography/tandem mass spectrometry. Rapid

Communications in Mass Spectrometry, 25, 2603-2610.

17.LaCroix, C., Saussereau, E. (2012) Fast liquid chromatography/tandem mass spectrometry determination of cannabinoids in micro volume blood samples after dabsyl derivatization. Journal of Chromatography B, 905, 85-95.

18.Scientific Working Group for Forensic Toxicology (2013) Standard Practices for Method Validation in Forensic Toxicology. Journal of Analytical Toxicology, 37, 452-474.

19.Logan, B., Lowric, K., Turri, J., Yeakel, J., Limoges, J., Miles, A., Scarneo, C., Kerrigan, S., Farrell, L. (2013) Recommendations for Toxicological Investigation of Drug-Impaired Driving and Motor Vehicle Fatalities. Journal of Analytical

Table 4.1. Selected ion transitions

Compound Precursor Quantifier Qualifier

THC-COOH 343 299.06 244.99

THC-COOH d3 346 301.89 N/A

11-OH-THC 331 312.99 193.10

THC 315 192.98 123.03

Table 4.2. The standard deviation of the y-intercept (Sy), average slope (n=5) (Avgm), limit of detection (LOD) in ng/mL, limit of quantitation (LOQ) in ng/mL, and the average coefficient of determination (R2_{) (n = 5).}

Sy Avgm LOD LOQ Avg R2

THC 0.001 0.014 0.22 0.67 0.9984

11-OH-THC 0.006 0.074 0.25 0.75 0.9980

Table 4.3. The accuracy, within-run precision, and between-run precision calculated as described above. Accuracy 2 ng/mL 10 ng/mL 25 ng/mL THC 4.7% 0.7% -0.2% 11-OH-THC -1.3% -2.8% 0.02% THC-COOH 3.6% 0.7% -0.4% Within-Run Precision 2 ng/mL 10 ng/mL 25 ng/mL THC 6.8% 4.6% 2.8% 11-OH-THC 5.0% 3.5% 1.8% THC-COOH 6.4% 4.5% 3.0% Between-Run Precision 2 ng/mL 10 ng/mL 25 ng/mL THC 7.5% 5.6% 2.9% 11-OH-THC 6.6% 6.0% 3.5% THC-COOH 7.0% 4.4% 3.1%

Table 4.4. Extraction efficiency, precipitation loss and matrix effects percentages for THC, 11-OH-THC, and THC-COOH at multiple concentrations.

Extraction Efficiency Precipitation Loss Matrix Effects